agentleFS
VepricovVepricov/claude-brainlab/skills/code-ingest/SKILL.md

code-ingest

Skill35 starsChanged 2 months agoMITView on GitHub

Adds a code repository to the Obsidian code library with its purpose, usage, and links to research projects. Use when the user hands over a repo URL or path and says 'add to the code library', 'разбери репозиторий', or starts a project that uses a repo missing from Code/.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. About 0 are common, 3 are distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Before heavily editing an unfamiliar repo (build the map first).

    line 17
  2. Very specific:

    Update Code/README.md (the library index) with a one-line entry for the new repo.

    line 37
  3. Very specific:

    Get the source: Prefer reading a local clone or the server copy if one exists (ask / check the project card).

    line 40
  4. Distinctive:

    Keep each note focused; 200-400 lines max, split if larger.

    line 105
  5. Distinctive:

    Use Obsidian wikilinks between the overview and module notes.

    line 107
  6. Very specific:

    After ingesting, remind the user to add a ## Код section to the relevant project hub card (see [[code-library]]).

    line 108

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

code-ingest — add a repository to the Obsidian Code library

Turns a repo into a map you can read in 2 minutes and instantly recall how the code is organized and where to look to change something — without duplicating the source. The code itself stays in git / on the server; these notes are the index.

When to use

  • The user hands you a repo URL or path and says "add to the code library" / "разбери репо".
  • A project starts depending on external code that has no Code/<repo>/ folder yet.
  • Before heavily editing an unfamiliar repo (build the map first).

When NOT to use

  • The repo is already in Code/<repo>/ and unchanged → just read it. Re-run only to refresh after large upstream changes.
  • For per-project edits/diffs → those go in the project hub card ## Код section, not here (see [[code-library]]).

Where it lives

Vault (the lab tree): ${OBSIDIAN_VAULT}/personal/code-library/ One folder per repo: personal/code-library/<repo-slug>/. The slug is the repo name, kebab-case.

Output structure (always)

Code/<repo>/
  <repo>.md          ← CANONICAL OVERVIEW. Always this exact name. Read this first.
  architecture.md    ← entrypoint, control/data flow, how a run executes end-to-end
  <module>.md        ← one note per major module/package (group tiny files)
  ...
  • The overview <repo>.md is mandatory and always named after the folder, so any agent knows which file to open first.
  • Split into module notes when the repo is large; for a tiny repo a single <repo>.md is fine.
  • Update Code/README.md (the library index) with a one-line entry for the new repo.

Procedure

  1. Get the source. Prefer reading a local clone or the server copy if one exists (ask / check the project card). Otherwise fetch file listing + key files from GitHub (raw URLs / gh). Note the canonical git URL and default branch.
  2. Map the layout. List top-level dirs and the role of each. Identify: entrypoint(s), config system, the main loop, the factory/registry or dispatch points (where names map to classes), data path, outputs/checkpoints.
  3. Trace one full run end-to-end (what main does, in order) — this becomes architecture.md.
  4. Write module notes using the template below. For each module: what it's for, key files with GitHub permalinks and line anchors, key classes/functions, and an explicit "where to edit to change X" list. Do not paste code bodies — cite path:line + a one-line description.
  5. Write the overview <repo>.md using its template.
  6. Index it in Code/README.md.
  7. Tell the user what was created and link the overview.

Overview template (<repo>.md)

---
tags: [code, <repo-slug>]
repo: <git url>
default_branch: <branch>
created: DD-MM-YYYY
---
# <repo> — code map

One-paragraph: what this repo is and what you use it for.

## At a glance
- **Entrypoint:** `path` (`func`)
- **Config:** how runs are parametrized (flags / yaml / hydra)
- **Run a job:** the minimal command
- **Data:** where data lives / how it's loaded
- **Outputs:** where checkpoints / logs / metrics go
- **Dispatch points:** where a name (optimizer/model/dataset) maps to a class

## Directory map
| path | role | note |
|------|------|------|
| ... | ... | [[architecture]] / [[<module>]] |

## Where to look to change X
- change the model arch → [[models]] (`...`)
- add an optimizer → [[optimizers]] (`...`)
- change the schedule → ...
- change data → ...

## Notes
- [[architecture]] · [[optimizers]] · [[models]] · [[data]] · ...

## Used by projects
- [[<направление>/<тема>/<работа>/README|<работа>]] — <fork/branch + what's changed>

Module note template (<module>.md)

# <module>

What this module is responsible for, in 2-3 sentences.

## Key files
- `path/to/file.py` — role. [GitHub](permalink)
  - `ClassOrFunc` (`file.py:LINE`) — what it does.

## How it fits the run
Where this is called from (link [[architecture]]), inputs/outputs.

## Where to edit to change X
- ... → `file.py:LINE`

Rules

  • Never copy large code blocks. Cite path:line + GitHub permalink + a sentence. The notes are a map, not a mirror.
  • Keep each note focused; 200-400 lines max, split if larger.
  • Pin permalinks to the analyzed commit/branch so line anchors stay valid; note the branch.
  • Use Obsidian wikilinks between the overview and module notes.
  • After ingesting, remind the user to add a ## Код section to the relevant project hub card (see [[code-library]]).

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.